Retinal optical coherence tomography (OCT) is increasingly used for quantifying neuroaxonal damage in diseases of the central nervous system such as multiple sclerosis. High-quality OCT images are essential for accurate intraretinal segmentation and for correct quantification of retinal thickness changes. The quality of OCT images depends largely on the operator and patient compliance. Quality evaluation is time-consuming, and current OCT image quality criteria depend on the experience of the grader and are therefore subjective. The automatic graderindependent real-time feedback system for quality evaluation of retinal OCT images, AQuA, was developed to standardize quality evaluation and data accuracy. It classifies by signal quality, anatomical completeness and segmentation plausibility and has been validated by experienced graders. However, it is currently limited to OCT scans taken with one device from a single vendor. The aim of this work is to improve the capability of the AQuA quality classifier to generalize to new data, by developing a convolutional neural network (CNN), AQuANet. Moreover, this CNN may serve as a basic quality classifier, that can be adapted to specific problems by transfer learning. AQuANet is trained on A-Scan batches with quality labels automatically obtained with AQuA. Thus, a large set of training data of about 13000 A-Scan batches could be used, leading to an accuracy of 99.53%.
The use of modern medical equipment in crisis and war zones for emergency medical teams (EMT) of the World Health Organization is an important factor for fast and efficient humanitarian aid. A reliable vital parameter monitoring is fundamental in mobile hospitals. Currently, the maintenance of medical devices in structurally weak areas is difficult due to the company’s proprietary standards. Rough environmental influences such as dust, moisture, heat or shocks can lead to dysfunktion and long-lasting failure of instrumentation. Pulse oximetry and blood pressure measurements are particularly susceptible. We developed an open source vital parameter monitoring system for use under adverse conditions and structurally weak areas. Blood oxygen levels, heart rate, blood pressure and electrocardiograms are recorded and transferred to decentralized displays. The main focus is on reliability and robustness of various optical sensors for pulse oximetry, the repair capability of the system also for non-technical personnel and the availability of individual standard components. Therefore we implemented a monitoring system basing on individual microcontrollers for each vital parameter. Different optical sensors for measurement in transmission and reflection were tested at suitable body sites with near-surface arteries. In combination with the electrocardiogram, evaluation of the pulse transit time enables continuous blood pressure measurements. A specially developed optical reflective sensor allows reliable measurement of blood oxygen level. For extended blood pressure measurements, the pulsetransit-time method (PTT) was implemented and enables a continuous monitoring. Even in emergencies, the trend in blood pressure can be monitored with PTT without prior calibration. The reliability was investigated.
One approach for a detailed understanding of dynamical cellular processes during drug delivery is the use of functionalized biocompatible nanoparticles and fluorescent markers. An appropriate imaging system has to detect these moving particles so as whole cell volumes in real time with high lateral resolution in a range of a few 100 nm. In a previous study Extended depth-of-field microscopy (EDF-microscopy) has been applied to fluorescent beads and tradiscantia stamen hair cells and the concept of real-time imaging has been proved in different microscopic modes. In principle a phase retardation system like a programmable space light modulator or a static waveplate is incorporated in the light path and modulates the wavefront of light. Hence the focal ellipsoid is smeared out and images seem to be blurred in a first step. An image restoration by deconvolution using the known point-spread-function (PSF) of the optical system is necessary to achieve sharp microscopic images of an extended depth-of-field. This work is focused on the investigation and optimization of deconvolution algorithms to solve this restoration problem satisfactorily. This inverse problem is challenging due to presence of Poisson distributed noise and Gaussian noise, and since the PSF used for deconvolution exactly fits in just one plane within the object. We use non-linear Total Variation based image restoration techniques, where different types of noise can be treated properly. Various algorithms are evaluated for artificially generated 3D images as well as for fluorescence measurements of BPAE cells.
Extended depth-of-field (EDF) microscopy is a well-investigated and very simple method to obtain projection images with an extended depth of focus. Despite its advantages of being a real-time method applicable to any microscopic mode with high lateral resolution that can be simply realized by extending a commercial microscope, the lack of z-correlation is still a problem. In this work we present a combined technique of EDF and stereomicroscopy. By cross-correlation depth information is obtained. Finally, 3D images are reconstructed for best phase masks and simulation results are evaluated experimentally.
Pupil phase masks for enhanced depth of field microscopy were investigated by using a spatial light modulator. The
phase masks were evaluated with simulations in terms of the mean square error between in-focus and out-of-focus point
spread functions. The resulting best-performing phase masks were tested for fluorosphere samples using a microscope
add-on containing the SLM. First, z-stacks of fixed fluorospheres in an agarose medium were recorded in order to
measure the extended depth of field. The same measurements were also performed on fluorospheres subjected to
Brownian motion in an aqueous solution. The results show that with deconvolution and appropriate filtering it is possible
to obtain sharp fluorosphere images with an extended depth of field of at least 10 μm.
Real-time visualization of live-cell dynamic processes has been realized in differential interference contrast (DIC)
microscopy, with an extended-depth-of-focus (EDF) increase of about one order of magnitude. In addition, the
diffraction-limited lateral resolution of the microscope is preserved. Experimentally, a custom-designed waveplate
inserted in the optical path of a microscope causes feature information, from within the entire 3D specimen volume, to be
uniformly encoded into a single CCD image in a way that, after processing, defocus blur artifacts are removed. The
result is that extended-depth feature information can be visualized at video rates during live-cell dynamics investigations
because there is no longer the need to acquire multi-focus image stacks at each time point. Retrieving the encoded
extended-depth information requires specialized digital image processing techniques. This work concentrates on digital
filter design for the reconstruction of the waveplate-encoded images. As a measure of filter quality, the signal-to-noise
ratio (SNR), the modulation transfer function and the least mean square values are evaluated. Obtaining a high SNR and
a lateral resolution comparable to those in conventional single-focus-plane microscopy images at the same time is a
challenging goal in EDF microscopy. Filters are created in the frequency domain on the basis of the measured
waveplate-encoded point spread functions. Results show that it is possible to produce video-rate, extended-depth-offocus
images that have low noise levels and diffraction-limited resolution. This is illustrated by movies of fluorescent
beads and of cytoplasmic streaming in live stamen hair cells from the spiderwort plant, Tradescantia, using extendeddepth